Big Tech Is Going After Health Care

The way you fix the existing system, is by not trying to change it. You replace it instead. You build new, lighter, more nimble approaches to old problems, rather than trying to convince existing power structures to change. That's where any gains will come from by big tech entering the fray, if there end up being any.

Trying to change something like the hyper regulated, hyper stagnant, special-interest dominated US healthcare system, is a folly of an idea. You kill it instead by routing around it with new solutions wherever possible.

Simple example: the rapidly growing market for direct primary care subscription services [1]. Pair that up with guaranteed catastrophic coverage for all by the US Govt or state governments, and the US healthcare system as we know it is dead due to a simple route around that would dramatically lower costs. It's far easier to implode a failed bureacracy by replacing it in that manner, than by trying to change the existing bureacracy. You see that in everything from old media (which held a cartel stranglehold on broadcast access, the Internet routed around them), to the taxi medallion cartels (which Uber & Lyft murdered by routing around them instead of trying to convince them to change).

[1] http://www.businessinsider.com/direct-primary-care-a-no-insu...

It is possible that some of the noise (non-medically relevant changes) could be filtered algorithmically, and conversely conditions might be able to be picked up that might not be noticed by a normal provider. Physicians use those tools to diagnose disease all the time, that's why there is a thermometer, scale, and stethoscope in every doctor's office in the world, I don't understand how these tools aren't useful.

Your comment is a great example of the disconnect between people in this space and people not in this space. Do you actually think anyone in this space would believe that weight spikes are tied to CHF in all cases? You don't diagnose CHF by weight. It's something you trend in known patients.

Comments like this are one reason why doctors give zero shits about the potential of computing. Explaining even the basics of a narrow area of human physiology to some overly arrogant IT guy is maddening (arrogant is fine, for Pete's sake, it's not like doctors aren't arrogant). Yes, the doctors can be blamed for conflating computer scientists with the IT guy who came to replace their mouse. But the IT guy and the CS guy (always a guy) can both equally be blamed for their unholy arrogance. How many lives did you save today? Zero? Ok, at least I didn't spill any data (that I know of).

This can go round, and round, and round.

You have no idea how far down the problem goes. I know this because no one knows how far the problem goes. We sequence hundreds of thousands if not millions of genomes, and still we don't know. Do you really think we spend a decade in training and come out thinking everything is simple?!

Now, on the flip side, doctors also don't even know how to frame their problems, in no small part because they're only required to take 2 semesters of calculus. And then most of them punch out of math as fast as possible.

Computer scientists should be going to biology conferences. Go to ASCO. Pick a medical specialty, they have multiple conferences a year, I garauntee it. There's one or 10 in your city. If you want middle ground, look at microbiology, immunology, molecular biology. They use a fair number of quantitative methods (sequencing, mass spec, flow cytometry, etc).

Aerospace also has high compliance hurdles. But rockets are extremely phallic. Even urology can't beat aerospace on that axis.

And yes, having been in this space for a while, the fax is a great example, a loophole for lawyers, who still do lots of paperwork by fax, because there's case law to support the practice (not the security mind you, just the practice).

Neither the lawyers nor the doctors understand math. Between the two, the lawyers at least took a formal logic course.

Doctors think most of the software they use is just magic, if they think of it at all. The only other association with computers that they have is the IT guy who comes to replace their tower every other year.

FWIW, healthcare compliance is a solved problem (disclaimer: I work for the company that solved it - Datica.com).

The larger problem is that young companies don't understand what it takes to be compliant, nor the resources to do it themselves. Remember, compliance is all about proving what you say you do. If you can prove it, you're golden.

So it's not that SMS or email isn't HIPAA compliant, it's that the work, controls, and monitoring you'd need to put in place to prove compliance is a massive burden. This is where risk management comes into play. If risk isn't a core tenant of your offering, you're better off building the preferred UX (like using email over fax) and assuming no risk.